Post Go back to editing

ADuM764X "glitch filter?"

The data sheet for ADuM764X mentions a "channel-by-channel glitch filter" but I find no discussion of how this works or what effects it has.  Is there a description somewhere?  Do your other similar parts that do not necessarily mention the "glitch filter" perhaps have this feature, or definitely not?

By the way, your responses to a couple of the questioners here about whether your isolators properly transfer DC levels to the outputs are a bit confusing.  The data sheet can be a bit confusing on this.  For the ADuM764X, under "DC CORRECTNESS" on p. 17, the text says "If the decoder receives no internal pulses for more than approximately 5us, the input side is assumed to be unpowered or nonfunctional..."  I think the confusion is that "internal pulses" refers to pulses generated by the isolator itself, NOT pulses the user feeds to the inputs; they are like refresh pulses in a memory, and DO NOT change the output state, just refresh it if it has not been changed by an input level change.  So I think that DC levels to the inputs are definitely passed through unchanged if both input and output power supplies remain active.

  • Hi Jeffbenj:

    this glitch filter exists on every digital isolators. its function is to filter high frequency noises on input. those pulses whose width is less than this filter will not be passed to the output side.

    about the transfer on dc level signal, your understanding is right. 

  • I'd add that some digital isolators do not include a glitch filter, such as the ADuM140 series which use OOK signaling. The glitch filter adds directly to the propagation delay, so it may not be utilized in the chip design when the shortest delay is desired. This is also true of the ADN4650//51/52 600 Mbps LVDS isolator family.